论文部分内容阅读
基于快速扩展随机树(rapidly exploring random tree,RRT)的运动规划算法,通过随机采样的方式探索未知任务空间,具有概率完备性和较高的计算效率.该类算法在应用于无人机运动规划时必须对飞行距离、过程安全性和航路平滑度进一步优化.针对这一问题,首先对威胁环境、无人机运动学性能和探测能力建模,然后根据飞行特征设计了随机采样、威胁规避、路径可跟踪性以及全局与局部平滑性等优化策略,并构建快速平滑收敛RRT(quick and smooth convergence RRT,QS-RRT),最后以此为基础分别提出了面向已知和未知任务空间的无人机运动规划算法.仿真结果表明,该算法能够在保证飞行路径收敛性、安全性及其规划效率的基础上,有效缩短飞行距离,改善航路的可跟踪性和平滑度,增强在实际飞行过程中的可操作性.此外,该算法还易于在航路优化效果和规划效率之间权衡,增强了对不同规划任务需求的适应性.
Based on the motion planning algorithm of rapidly exploring random tree (RRT), the unknown task space is explored by random sampling, which has the probability completeness and high computational efficiency. This kind of algorithm is applied to the UAV motion planning We must further optimize the flight distance, process safety and route smoothness.To solve this problem, we first model the threat environment, UAV kinematics and exploration capability, then design random sampling, threat avoidance, Path traceability, global and local smoothness, and construct fast and smooth convergence RRT (QS-RRT). Based on this, we propose the unmanned algorithm for unknown and unknown task space respectively Machine motion planning algorithm.The simulation results show that this algorithm can effectively shorten the flight distance, improve the traceability and smoothness of the route, and enhance the efficiency of the flight in the actual flight process, while ensuring the convergence, safety and planning efficiency of the flight path, In addition, the algorithm is also easy to trade-off between route optimization and planning efficiency, Adaptation planning needs of the mission.